Integrating Technology in the Foreign Language Classroom

Scanning images
 General principals

In today's world of digital cameras and Google images, teachers have less and less need of photo scanners. Sometime, however, the ability to scan an image in order to include it in a presentation using authentic materials can be quite useful.

Scanning a photo on a flatbed scanner involves a number of basic steps that are essentially the same no matter what particular scanner and software are used.  As light is shone on the photo, photo-electric cells capture the red, green, and blue light (RGB) reflected from individual points of the image.  Each point is a pixel (picture element), and the resolution of the image is measured in pixels per inch.  The total number of pixels in a rectangular image can be found by multiplying the horizontal width in pixels by the vertical height in pixels. So an image that is 600x800 has 600 dots along one side and 800 dots along another. If it was scanned at 120 dpi, that means that one inch of the original photo was converted to 120 dots in the scanned image.

Every pixel in a color image can be represented by three numbers indicating the brightness of the red, green, and blue components of the color, and in fact, each pixel on a computer screen (or your TV) is made of three tiny dots, one red, one green, one blue.  Our eye then interprets these as the millions of colors that we perceive. The way colors are mixed to fool our eye into seeing nuances of color or perceiving a higher resolution than is actually present can be mathematically quite complex, but it is not essential to understand Floyd-Steinberg Diffusion Dithering and other such complexities to use them.

Different image formats store information about pixels and colors in different ways.  A GIF image, for example, can use a maximum of 256 different colors stored in an indexed color table.  A JPG image uses 256 different possible values for each of the three color elements (RGB) for a total of 256x256x256 =  16,777,216 different possible colors for each pixel. You may have seen colors represented in hexadecimal (base 16) FF080A = 255 red, 8 green, 10 blue. Image files take up large amounts of memory without some form of compression. GIF uses a compression scheme that works well for small numbers of colors without smooth transitions, but GIF compression causes no loss of image quality.  JPG allows the user to determine the degree of compression of the image, but although it can greatly compress a large number of colors and can handle smooth transitions, JPG compression is lossy (the more the image is compressed, the more information / quality is lost). Indeed, any time an image is converted from one format or size to another, some information is likely to be lost.

The quality of an image will depend on the resolution (the number of pixels used), the color depth (the number of possible color values for each pixel), the effectiveness with which the colors are used to represent the range of brightness and contrast in the picture, and any loss of quality caused by the compression of the image. For these reasons, the two keys to successful scanning are to scan the image at the largest resolution (size) that you will need and to capture with a brightness and contrast that look best to your eye.

The basic ideas behind scanning are the same from one scanner to the next. We will look at a few different examples. First we'll be using the Microtek E6 with ScanWizard software.

ScanWizard is launched from within an image editing application such as Photoshop. Select File/Acquire. Twain is the name of a common scanner interface found in an Acquire menu, though many of today's scanners will be found listed by name in an Acquire menu. In the Preview Window, set up the scanning mode.  It is important to verify that the correct Scan Material, Type, and Resolution settings are in place.  This will yield the best results for color, contrast, and final size.  For a color photo that you wish to display on a computer screen, the settings for reflective material, millions of colors, 72 dpi should work well.

    Check to be sure that Image Enhancement Tools (any tools or settings that modify the quality of your scan) are reset or set as you desire.  You probably don't want to use the same adjustments that someone else may have used previously on a different photo. For typical color photo scans in ScanWizard, do not use Color Correction and Auto at the same time.  It is probably better to select Color Correction and make manual adjustments for brightness, contrast, shadows, highlights, etc. because color is usually harder to adjust.  

Place your original at the front of the scanner, upside down. On some scanners the edge may be slightly cut off, so you might want to leave a small space.  Be sure that the original is not tilted.  Press Preview.  In the preview image, check that the photo is straight and select the area to be scanned.  If adjustments are needed to improve image quality, you may select Image Enhancement Tools or use photoshop to make adjustments later if the problems are not to severe.

A second example is the HP ScanJet 5p scanner.  It is launched also from within the File / Acquire menu in Photoshop, Paint Shop Pro, GIMP, or any other image editor.

The process of scanning a photo will follow the same procedure that was used for the Microtek ScanWizard software.  First check that the software is set for color pictures, then click Scan to get a preview image.

The marquee tool will let us draw a rectangle to select the area for the final scan.  The Options menu is needed to set the resolution (image size) of the output.  After the options have been set, clicking Accept will begin the final scan.

The ruler icon will allow us to change the resolution of the scanned image before completing the scanning process.

HP ScanJet 3500c

Photoshop

The most common adjustments are:
• Brightness/Contrast/Exposure time - Increasing brightness tends to decrease contrast so if you change one, you may need to adjust the other also. This can work if the overall brightness is not quite right, not when only a part of the image needs adjustment.
  
  
  
• Levels (Shadow & Highlight) - This tool works best when dark areas are too dark or light areas are too light or when adjustments can be spread out broadly in the light or dark bands.  Your objective, as when developing a photograph, is to make the best use of the contrast range of the medium by using all the colors from full black to full white.  Look at the histogram (a graphic representation of brightness distributions), move the black and white triangles to identify darkest and lightest colors, move the gray triangle to spread out mid-tones.


• Curve tool - This tool works well for bringing out details in the mid-tones of a photo.  It allows more precise control of a portion of the brightness range than the levels tool but can be harder to use. I prefer sticking with Levels.   The curve can be shaped in any way, and the slope of the curve will define the relation between input and output pixel intensities or brightness.  This is a powerful tool for precise adjustments of contrast in specific areas of a photo.


• Image size controls the number of pixels (picture elements, dots) in the size of the image.  Resolution can be confusing because it can have two very different results depending on the output device. In a printed image, the resolution may control how sharp the picture appears by putting more information in a given area of the image. On a computer screen with a fixed display resolution, an increase in image size / resolution cannot change the way the screen works; it will only make the image larger, using more screen area to display the additional information. Thus resolution is related to scaling or the size of the final image as related to the size of the original.  Doubling the resolution will make the final image twice as wide and twice as tall, but the image file will be four times as large (w/o compression).
  
  
  

Saving images:

Once an image is scanned and prepared, it must be saved as a file.  Files can be in GIF, JPG, PNG, or PICT format (among others).     Which to use?

• GIF - 256 colors, good compression, good for straight lines, few colors sharp contrasts.
• JPG - millions of colors, excellent compression, compression is adjustable but lossy (more compression = loss of image quality), good for photos with lots of colors, smooth color transitions.
• PNG - can be thought of as a compromise between GIF and JPG. It is intended for the WWW, but is not yet in very common use.
• PSD (PICT) - Photoshop default. No loss of quality on multiple saves. Best for a working copy that may need to be saved and reopened repeatedly. But the file size is very large and this will not work in a web page without conversion to GIF, JPG, or PNG. .

Scanning Guidelines: Simple Steps to Good Colour Scanning - by Jane Fritz
CAMDUS KNOWLEDGEWORKS - Digital Art Support